PyDESeq2 is a python implementation of the DESeq2 method [1] for differential expression analysis (DEA) with bulk RNA-seq data, originally in R. It aims to facilitate DEA experiments for python users.

As PyDESeq2 is a re-implementation of DESeq2 from scratch, you may experience some differences in terms of retrieved values or available features.

Currently, available features broadly correspond to the default settings of DESeq2 (v1.34.0) for single-factor and multi-factor analysis (with categorical or continuous factors) using Wald tests. We plan to implement more in the future. In case there is a feature you would particularly like to be implemented, feel free to open an issue.

Table of Contents

• PyDESeq2
  • Table of Contents
  • Installation
    • Requirements
  • Getting started
    • Documentation
    • Data
  • Contributing
    • 1 - Download the repository
    • 2 - Create a conda environment
  • Development roadmap
  • Citing this work
  • References
  • License

Installation

PyPI

PyDESeq2 can be installed from PyPI using pip:

pip install pydeseq2

We recommend installing within a conda environment:

conda create -n pydeseq2
conda activate pydeseq2
conda install pip
pip install pydeseq2

Bioconda

PyDESeq2 can also be installed from Bioconda with conda:

conda install -c bioconda pydeseq2

If you're interested in contributing or want access to the development version, please see the contributing section.

Requirements

The list of package version requirements is available in setup.py.

For reference, the code is being tested in a github workflow (CI) with python 3.9 to 3.11 and the following package versions:

- anndata 0.8.0
- numpy 1.23.0
- pandas 1.4.3
- scikit-learn 1.1.1
- scipy 1.11.0

Please don't hesitate to open an issue in case you encounter any issue due to possible deprecations.

Getting started

The Getting Started section of the documentation contains downloadable examples on how to use PyDESeq2.

Documentation

The documentation is hosted here on ReadTheDocs. If you want to have the latest version of the documentation, you can build it from source. Please go to the dedicated README.md for information on how to do so.

Data

The quick start examples use synthetic data, provided in this repo (see datasets.)

The experiments described in the PyDESeq2 article rely on data from The Cancer Genome Atlas, which may be obtained from this portal.

Contributing

Please the Contributing section of the documentation to see how you can contribute to PyDESeq2.

1 - Download the repository

git clone https://github.com/owkin/PyDESeq2.git

2 - Create a conda environment

Run conda create -n pydeseq2 python=3.9 (or higher python version) to create the pydeseq2 environment and then activate it: conda activate pydeseq2.

cd to the root of the repo and run pip install -e ."[dev]" to install in developer mode.

Then, run pre-commit install.

The pre-commit tool will automatically run ruff, black, and mypy.

PyDESeq2 is a living project and any contributions are welcome! Feel free to open new PRs or issues.

Development Roadmap

Here are some of the features and improvements we plan to implement in the future:

• [x] Integration to the scverse ecosystem:
  • [x] Refactoring to use the AnnData data structure
  • [x] Submitting a PR to be listed as an scverse ecosystem package
• [x] Variance-stabilizing transformation
• [ ] Improving multi-factor analysis:
  • [x] Allowing n-level factors
  • [x] Support for continuous covariates
  • [ ] Implementing interaction terms

Citing this work

@article{muzellec2023pydeseq2,
  title={PyDESeq2: a python package for bulk RNA-seq differential expression analysis},
  author={Muzellec, Boris and Telenczuk, Maria and Cabeli, Vincent and Andreux, Mathieu},
  year={2023},
  doi = {10.1093/bioinformatics/btad547},
  journal={Bioinformatics},
}

References

[1] Love, M. I., Huber, W., & Anders, S. (2014). "Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2." Genome biology, 15(12), 1-21.

[2] Zhu, A., Ibrahim, J. G., & Love, M. I. (2019). "Heavy-tailed prior distributions for sequence count data: removing the noise and preserving large differences." Bioinformatics, 35(12), 2084-2092.

License

PyDESeq2 is released under an MIT license.